A common technique for visualizing data, such as records or logs, is through a workflow diagram or flowchart, which can be abstractly represented as a graph. The data corresponding with the sequential actions of an individual case can be represented as paths on the graph. The paths on the graph can be condensed or compressed for analyzing characteristics of the graph. Conventionally, compressed paths are represented in the transitions (i.e., edges) of the graph by varying the thickness of the graph's transitions to reflect their frequency of occurrence in the dataset. This is helpful in judging the importance of individual steps in a session and identifying hot spots in the graph. However, properties of overall paths on the graph may be lost. More specifically, conventionally compressed paths of a graph may lose too much information to support comparison of certain actions.